SC14: Jack, you are well known for many pioneering contributions over a long career in HPC. Chairing the Technical Program is a huge commitment of time that you don’t have to make at this point in your career. Why did you agree to take on this challenge?
Jack Dongarra: The SC Conference is something special to me. I have attended all of the SC Conferences; in fact I helped start the series with George Michaels, Bill Buzbee and others. Over the years I have chaired many parts of the Technical Program, but never had a chance to chair the whole Technical Program. SC plays an important role in the high-performance community. It is through the SC Conference that HPC practitioners get an overview of the field, get to showcase our important work, and network with the community. There is really no other venue like it. I like to say it’s the supercomputing homecoming meeting with ten thousand of my closest friends. This year I’m also lucky to have the chance to work with Michela Taufer from the University of Delaware as my Co-Chair, and Bruce Loftis from the University of Colorado, Boulder as my Chair Assistant.
SC14: Each year SC builds on its core Technical Program by adding new elements that showcase changes in the field. What’s new in the SC Tech Program this year?
Dongarra: The SC Technical Program is highly competitive and one of the broadest of any HPC conference. Traditionally, the Program includes invited talks, panels, research papers, tutorials, workshops, posters, and Birds-of-a-Feather sessions. SC14 will also be launching new initiatives focused on big-data and analytics, as well as on innovative new technologies in HPC. Alongside the traditional program elements that our community has come to rely upon, SC14 will continue to be HPC’s most informative, exciting, and stimulating event of the year.
We’re especially excited to be highlighting data analytics this year as we introduce a new format for the Visualization and Data Analytics Showcase Program at SC14, which provides a forum for the year’s most instrumental videos in HPC. Six finalists will compete for the Best Visualization Award, and each finalist will present his or her movie in a 15-minute presentation during a dedicated session at SC14. Submissions will be judged based on how well the analysis illuminates the underlying science, by the quality of the end result, and for innovations in the process used for creating the movie. The idea is to describe the scientific story conveyed by the movie, how the visualization helps scientific discovery, and the “state-of-the-practice” information behind making the movie.
In addition, we are continuing the Test-of-Time award that was started at SC13. The Test-of-Time award recognizes an outstanding paper, first presented at a past SC Conference, which has deeply influenced the HPC discipline. The award is also an incentive for researchers and students to send their best work to SC, and a tool to understand why and how results last in the HPC discipline. Any SC paper that is at least 10 years old is eligible, and selection is made by a special committee. Last year the winner was William Pugh from the University of Maryland for “The Omega Test: a fast and practical integer programming algorithm for dependence analysis,” published in the proceedings of Supercomputing’91.
We are also happy to announce that the Doctoral Showcase Program is back in the Technical Program at SC14. This year the program wants to be the place where prospective employers in academia, research laboratories and industry can recruit the next-generation of HPC experts.
SC14: Last year’s Tech Program highlighted a commitment to extending peer review to all SC Tech Program elements. Is that something that you are continuing? Why is that important for attendees?
Dongarra: Peer review is commonly accepted as an essential part of any scientific publication. Every component of the SC Technical Program undergoes peer review by the community of experts. In many cases four referees participate in the reviewing process, and for SC this involves over 300 people across the program elements. The authors of submitted papers will receive the reviews and will have an opportunity for rebuttal during which factual errors in the reviews can be addressed, and for answering specific questions posed by reviewers. The rebuttals will be read by the referees and factored into the discussion leading up to the acceptance decisions made by the Technical Papers Committee. The whole Technical Papers Committee will meet in June to discuss each paper, the reviews and rebuttals. All members of the Committee are expected to attend in person and participate in the reviewing process. It’s an impressive collection of people; really, the technical leadership in HPC has been solicited to participate.
In general the peer review process helps to maintain standards of quality, improve performance, and provide credibility. I have seen the quality of accepted papers at SC steadily increase over the years, and the competition has been also increasing. So that getting a paper accepted SC has value, especially for someone just starting their career.
SC14: Over your career you’ve seen several technology cycles in HPC. What is your sense for where we are in supercomputing now? Do you see us in a trough tweaking commodity technologies to get a few extra percentage points in improvement, or are we ripe for a fundamental shift in computing that will open new opportunities for us?
Dongarra: This is one of the most exciting times in my career because of the challenges we face with the coming of exascale computing. We face a host of challenges with the potential to reshape our community: from hardware challenges with energy-efficiency, memory technologies, interconnect technologies; software challenges with scalable systems, programming systems, and data management; algorithmic challenges with new methods at exascale; resilience and correctness; and scientific productivity.
Exascale computing will provide capability benefits to a broad range of industries, including energy, pharmaceutical, aircraft, automobile, entertainment, and others. More powerful computing capability will allow these diverse industries to more quickly engineer superior new products that could improve a nation’s competitiveness. In addition, there are considerable flow-down benefits that will result from meeting both the hardware and software high-performance computing challenges. These would include enhancements to smaller computer systems and many types of consumer electronics, from smartphones to cameras.